D&D 5E Is there a general theory of party construction?

[payn]

I like players having a lot of options to round out a party and prefer versatile characters. I like class systems, but prefer ones that let you choose which direction to lean in. I don't like when classes are hard coded into a role. So, the ambiguity in 5E doesn't bother me.

 

[jmartkdr2]

The only role you need is a front line - because if everyone's in the back line, that's the new front line. (or: if you don't pick a frontliner, the enemy will).

Other than that, for 5e there's no real "needs." It's better to focus on what your pc is intended to do and make sure you're good at that.

 

[tiornys]

The only role you need is a front line - because if everyone's in the back line, that's the new front line. (or: if you don't pick a frontliner, the enemy will).

I sort of agree with this, but probably not in the way you meant it. In 5E I think every character should be capable of functioning in melee, but no character should have "get into melee" as their primary combat plan. Having a "frontliner" in the classic sense doesn't make sense in 5E; the risks of being in melee outweigh the benefits and melee doesn't do enough on its own to present any real control against being bypassed. On top of that, melee doesn't do enough to hinder spellcasting and ranged characters have strong outs to getting stuck in melee. Conclusion: everyone should plan to be ranged and/or spellcasting as their primary mode of combat, and yes, that means your "back line" is also your "front line" (but really your front line is some sort of crowd control measure e.g. Sleep, Grease, Web, Plant Growth, Spike Growth, summoned creatures, Spirit Guardians, Hypnotic Pattern, Sleet Storm, etc. etc.).

 

[ECMO3]

This is a perennial question, but I was wondering if anyone ever tried to concisely say what bases needed to be covered in a D&D party. 4th edition had the system of striker, controller, leader (support really), and defender, but since then it seems somewhat unclear. We have the sense a party of 3 sorcerers and 2 wizards would be unbalanced, but apart from that...? Seems like it would make it easier for, say, a player joining an existing group of 4 to know what the party needed, for instance.

In 5E the general theory is build what you want to play.

The only 2 things I would say:

1. Someone in the party should probably have thieves tools proficiency

2. It is generally fine if you build all characters of the same class, but if you have the exact same type characters you will run into a problem where there will be few useful magic items. For example 4 fighters in fine, but if you want to play 4 fighters in plate and swinging a greatsword now that magic studded leather armor and the magic sunblade you found are nearly useless.

 

[EzekielRaiden]

This will of course be intensely controversial because some folks view D&D in a maximally loosey-goosey, anything-goes, "it's just FUN stop trying to STUDY it, GOSH," but we can do quite a bit of analysis on this.

5e openly features three pillars: combat, exploration, and socialization. These have various expectations. Combat is generally the easiest to optimize for, and the least open (not NOT open, just less so) to unpredictable DM intrusions that can throw off analysis, thus it gets a lot of attention and effort. This is where much of the (often undeserved) bad reputation of optimization comes from. By knowing typical HP, AC, and saving throw values, common resistances and immunities, and the types and likelihoods of various creatures, it is possible to make fairly accurate claims across a broad swathe of plausible scenarios. E.g., players who wish to be at least decent at combat should generally be able to do at least an unmodified damage cantrip's worth of damage each round, or something more valuable but not directly comparable (e.g. locking down enemies so they cannot respond to the PC or their allies.)

For the combat pillar, we can generally say the following things:

• The best saves are Wis, Con, and Dex, generally (but not always) in that order. Characters should all have at least one of those saves Proficient simply by 5e's design, and if the party is lacking in proficient saves of any of these, that's a thing to consider. The other three saves CAN be bad if failed, but are dramatically rarer especially before high levels.
• As noted, cantrip's provide a solid baseline of expected damage per round. There will always be times where doing damage is not as valuable as some other goal. But in the generic, firebolt is effectively a baseline of DPR.
• Spells are very powerful in combat (as they are in every pillar) and usually a character is stronger if they can cast spells than if they can't. This increases significantly if the party does not actually follow the expected "about 2-3 short rests per day, and about 6-8 encounters, most of them combats, every day" pattern, which unfortunately most groups do not follow.
• If possible to do so without sacrificing a lot of damage, use Dexterity to attack rather than Strength. Dexterity boosts a much more powerful save, AC, the all-important Initiative, and a larger number of skills. Some classes cannot pull this off very effectively or provide incentives that make it worth considering Str instead of Dex, but again, this is a general rule not a special exception.
• Your three highest stats should be Dex, Con, and whatever your focus stat is for your class, e.g. Paladin/Bard/Sorcerer value Cha, Cleric/Druid/Monk value Wisdom, etc. There may be exceptions as noted (e.g. Paladins may prefer Str/Con/Cha), but overall this is a good pattern.
• Certain types of equipment are Just Better or Just Worse, e.g. the rapier is the best one-handed melee weapon for anyone who can wield it, because it is a d8 finesse weapon, as good as any Str weapon but usable with Dex. The trident, meanwhile, is strictly inferior to the spear: it is heavier, more expensive, and belongs to a more difficult category of weapons (spears are simple, tridents are martial), but otherwise has exactly the same mechanical properties. Why bother with it when the spear exists?
• Some effects or damage types are much less useful than others. Necrotic and poison, for example, are extremely common resistances or immunities, as many undead, celestial, and/or fiendish enemies have such resistance or immunity; having resistance to necrotic is pretty useful though. Fire is commonly resisted, but also readily available in a lot of spells, so it kind of balances out. Radiant is very rarely resisted, but hard to find as a damage type both offensively and defensively. Etc. Use these facts to shape your class and spell choices.

I could probably add more. Point being, we can do a lot to narrow down useful things most if not all groups should strive to have, and thus make a "theory of party composition" as you put it.

From my time thinking on the subject I would say every party of at least 4 people needs the following (some of which may be doubled up on a single char):

• At least one magical healer. Sadly, non-magical healing just cannot keep up with damage in 5e, and the resting rules don't help on that front.
• At least one full caster with the ability to cast rituals. Preferably two of distinct traditions (e.g. Cleric and Wizard, Druid and Bard, Bard and Wizard, etc.), but one will suffice.
• At least one character with high HP and good defenses (which may mean AC, damage resistances, or spells that prevent or mitigate damage).
• At least one (and preferably as many as possible) character with Perception proficiency. It is the most important skill in the game. It should not be ignored. Ideally at least one such character also has high Wis, but it's not essential.
• At least one character proficient in each of the "major" saving throws (Con/Dex/Wis). Preferably, there's at least someone proficient with any of the six, but that's harder to pull off with only 4 people.
• At least one character whose focus ability is Charisma and who has trained at least one Charisma skill. Faces are very important.
• At least one character whose focus ability is Dexterity and who has Stealth proficiency (and hopefully other Dex skills). Ideally they also have lockpicking tool proficiency, but that's not strictly necessary.
• The party employs a diversity of damage types. Unless you can be confident that the party won't deal with tons of fire-immune critters, you want to keep damage type diversity. Psychic, force, and radiant damage are generally good against most enemies, while necrotic and poison are unreliable. It's okay for individual characters to specialize so long as the group overall has decent coverage (easily achieved by most full casters).
• Optionally, one character with high Strength. Strength is a somewhat niche stat in 5e, good for specific things but not very broad. However, it's reasonably likely that situations where strength would be useful will pop up, and it's nice to not need spells to deal with those. It is not very useful to have more than one or two such characters in a party due to the risk of over-specialization.
• Similarly optional, either one character specialized in skill expertise (e.g. Rogue or Bard) or in Intelligence is useful. Intelligence almost exclusively affects knowledge skills unless you're a Wizard or Artificer, but those knowledge skills can be useful in the right contexts. (It also doesn't help that the difference between Investigation and Perception is extremely unclear in many groups.)

That covers most of what I can think of off the top of my head. I'm sure I've missed things, but this is fairly comprehensive nonetheless.

A "solid group" of four is definitely met by the classic Fighter/Rogue/Wizard/Cleric, but you could also do Moon Druid/Bard/Artificer/Paladin, or Barbarian/Rogue/Sorcerer (with Ritual Caster)/Land Druid, or a variety of other substitutions.

 

[AtomicPope]

Gygax covers this in his book "Rules Mastery" (or the other one) when he describes the development of D&D from being a miniatures based war game to an RPG. The Thief (skilled class) was the missing link. In a minis war game you had your medic (cleric), artillergy (wizard), and armored tank (fighter). Yes, Gygax does call the fighter a "tank". All of that is fine for a war game but D&D is much more than that. The first D&D adventure was breaking into a castle and fighting a dragon in a dungeon. From the beginning they saw that there was a need for two things: 1) every PC must have a vital role 2) a skill class is needed to fill a necessary role.

Over time the DMs were changing their ideas on what constituted a "campaign". Even Gygax got away from the dungeon crawls and with "Keep on the Borderlands" D&D started to look more like an open world RPG. Nowadays it's common to talk about "themed campaigns" where everyone plays a Wizard or Rogue, but that was being discussed in the earliest days of D&D. Gygax and company had their "Circle of Eight" which was just a bunch of Wizards getting stuff done.

The nature of the campaign will determine what is "needed". I'm currently running and Eberron Detective Noir that thematically resembles "Blade Runner" more than D&D or Steampunk. Warforged are more like Replicants (some look very human). Since there was no planned obsolescence, the Five Kingdoms passed a "Registration Act" whereby all Warforged must be registered, and their weaponry decommissioned, or be decommissioned. In this game no one is playing a Fighter, Wizard or Cleric. There's two Artificers, a Radiant Monk (who uses laser pistols), a Paladin/Warlock and a Changeling Rogue. This campaign doesn't have many "boss fights" since most of the work is skill based. The Paladin/Warlock has skill invocations and Disguise Self at will. This is a skill based campaign. The healing is minor and so are the number of fights.

One thing I do like about 5e is the Bounded Accuracy and Short Rests give both players and DMs a lot of leeway that didn't exist before. It does mean that Clerics are no longer a hard requirement for every campaign, and that's not a bad thing.

 

[Lanefan]

The only role you need is a front line - because if everyone's in the back line, that's the new front line. (or: if you don't pick a frontliner, the enemy will).

Yep.

Rule one for party construction in any edition is this: When in doubt, add more front-liners. You can never have enough.

Rule two is this: You can never have too many characters in a party. If you even think you might need more, go and get more, right now.

 

[Mind of tempest]

frontliner, healer, face/skill monkey.

 

[EzekielRaiden]

Yep.

Rule one for party construction in any edition is this: When in doubt, add more front-liners. You can never have enough.

Rule two is this: You can never have too many characters in a party. If you even think you might need more, go and get more, right now.

Meh. I know that in earlier editions, retainers and hirelings were hugely important, but it's an awful lot of hassle in every WotC version, with the possible exception of 4e. Both 3e and 5e have "NPC class" options that are supposed to be more streamlined, but they really aren't much different from a complete character with crappy class features. They're really not much faster to run than literally just adding additional characters, giving a ton of overhead for ever-more-marginal gain.

And I dunno about you, but the thought of running a combat in 3e or 5e with 7 or more PCs sounds nightmarish.

 

[EzekielRaiden]

Working off of my above post (now that I'm not writing it on a tiny phone screen), I can give a more precise accounting.

In terms of combat-related actions, you must have (or, for some of them, really really should have, but don't technically need) at least one of each of the following (one person may possess multiple elements):

1. Weathering sizable or sustained damage without going splat; the "armored tank" as Gygax put it
2. Focusing enemy attention where it is least harmful or most beneficial to the party
3. Dealing good damage to several creatures at the same time
4. Dealing damage from a distance
5. Healing (though in general this is for "post-combat" rather than during)
6. Dealing high spike damage to individual targets
7. Closing the distance with the enemy, often to deny any terrain advantages
8. Creating terrain advantages for allies or terrain disadvantages for enemies
9. Providing force-multiplier effects (other than terrain), or removing sources that weaken allies
10. Taking away enemy force-multiplier effects (other than terrain), or applying effects that weaken enemies

In 4e, these were organized into the four roles. Being a "Defender" class meant having baseline class features that supported 1, 2, and 7, and to a lesser extent 3, 8, and 10; it was almost always possible to spend resources (usually feats and/or items, but sometimes class powers as well) so that any class could provide some or all of those benefits, but not guaranteed. A Wizard, for example, might not be very good at personally taking hits, but if she were specialized in summoning or illusions, she could be creating extra targets on the battlefield that would cover 1 and 2 above without any extra costs to the party.

For 4e's other roles, Strikers usually hit most or all of 3, 4, 6, and 7, with some light splashes of 9; Leaders focused heavily on 5 and 9 but often picked up one or two of 1, 2, 3, 7, 8, and/or 10; and Controllers specialized in 3, 8, and 10, but often also offered 4 and 9 too. As you can see, there's a lot of overlap. E.g. 10 shows up in different ways under Defender, Leader, and Controller; 9 was a focus of Leaders but also appeared with Strikers and Controllers; 7 appears frequently for all roles except Controller, and that's only because Controllers usually focus on areas and multiple targets, while 7 is mostly a concern for melee-focused classes/builds. Etc.

We can do a similar analysis of social situations, though this is more loosey-goosey because (as has always been the case in D&D) there are fewer and lighter rules for social situations than for combat ones. Social interactions have both a broader scope and fewer defined elements, e.g. there is no equivalent of "HP" defined for social situations. (This would be part of why 4e proposed the Skill Challenge, even if the framework was very rough to begin with--with things left highly loosey-goosey, it's easy to fall into the trap of "mother-may-I" and harder to think critically about how one achieves success outside of "keep poking and prodding until the DM relents, rejects, or rolls.")

In social situations overall, e.g. covering as many potential situations as possible, the group will need at least one person who can:

1. Identify false statements made by others and the true motives of others
2. Make convincing false statements and conceal the PCs' true motives
3. Detect disguises and/or illusions and otherwise ascertain the true state of affairs
4. Create convincing disguises and/or illusions and otherwise conceal the true state of affairs
5. Speak eloquently about the party's intent, purpose, etc.
6. Coerce others into doing as the party desires, even if they would rather not (whether by threats, offers, oaths/values, etc.)
7. Stand up for the party's goals or desires in the face of persuasive force (e.g. resist the aforementioned tools of manipulation)
8. Act as translator or interpreter for situations where language is a barrier
9. Acquire and preserve important evidence and information for later use
10. Persuade or manipulate large groups of people all at once (via performance, subterfuge, what-have-you)

It's possible there are more that I'm overlooking, but this is reasonably comprehensive. Most of these elements are covered in one of three ways: having a relevant background characteristic (whether a formal in-the-rules Background feature, or something separate from character features, e.g. a backstory element), having Proficiency/Expertise in a skill and/or a high ability modifier for the associated ability of that skill, or using magic. Unfortunately, as with the above set, some of these elements in 5e pretty much require magic, whether as a hard requirement (you really can't make illusions unless you can cast spells) or a strong but still technically soft requirement (translation is vastly easier with magic, preserving many things such as corpses is nigh-impossible without magic, and while it is technically possible to detect magical things without detect magic, it's often damned hard.)

The other unfortunate side of this pillar is that it is often possible for one single person to be legitimately excellent at every single one of the above elements. The Bard class, for example, can theoretically fill every single one of those roles with high competence. A spellcasting-focused Lore Bard with high Cha and Wis can be borderline-unbeatable in nearly every category simultaneously. Note that I do not say this because I think the Bard is a problem, because I don't think that--I think it's merely illustrative. I say it because it reflects how minimally-developed the mechanics of socialization are in 5e (and 3e; I do not have enough experience with 2e or earlier to speak about them), and that making them even less developed does not (and I would argue cannot) solve that problem. Some of it, of course, is that low-level magic is often an extreme binary of either fantastically useful or totally worthless (e.g. disguise self is perfect so long as no one tries to touch you, and worthless if they do touch you; friends is either utterly pointless because it makes people hostile a mere minute later, or fantastically OP because you don't care if the target becomes hostile). But some of it is also just that "have good numbers in 4-5 skills" is something most characters can achieve, and likewise "be able to cast some illusion, enchantment, divination, and/or transmutation spells" is something nearly every full caster can do by the time everyone has their subclass (and even many non-full-casters can do it).

Most people have an intuitive sense that if a single character could do excellent single-target and area-of-effect damage, buff others, debuff enemies, control the battlefield, heal, take big hits, and work equally well at range or in melee, that that character would be overpowered and inappropriate. Even people who tend to treat "balance" as a stinky no-no word will speak out against something they perceive as capable of doing this. But as soon as you move away from combat, that intuitive sense disappears, and people seem perfectly content with having characters that can do literally everything in a pillar at high to maximum competence.